The resistance most leaders are trying to overcome isn’t resistance to AI. It’s resistance to being managed about it.
In June 2026, PagerDuty published a survey conducted by Wakefield Research among 1,250 office professionals in non-technical roles at companies with at least $500 million in revenue, across Australia, Japan, the UK, and the US. Two-thirds reported using AI tools at work that their employer hadn’t authorized. A third said they would hide their AI use specifically to avoid scrutiny from managers or leadership.
Read those two figures together and the standard framing collapses. The problem in most organizations is not that people won’t use AI. They’re using it enthusiastically, at their own expense, on personal accounts, at some professional risk. What they won’t do is route that use through the front door.
A separate global survey, Lenovo’s 2026 Work Reborn research covering 6,000 full-time employees at enterprise organizations, points at why: somewhere between a fifth and a third of workers use AI entirely outside IT’s oversight, and a substantial share report receiving no employer training at all — with many of those who do describe it as irregular or ineffective. The gap isn’t between willing and unwilling employees. It’s between people who were given tools, training, and permission, and people who were given a policy.
So the instinct to push harder — another mandate, another all-hands, another compliance dashboard — is aimed at a barrier that isn’t there, and it tends to reinforce the one that is.
Jonah Berger’s The Catalyst is built entirely around that distinction — between resistance you push through and barriers you remove. His argument inverts most of what leaders are taught about driving change: the way to move someone isn’t to add more pressure, it’s to find and release the parking brake. He organizes the barriers into five, under the acronym REDUCE — reactance, endowment, distance, uncertainty, and corroborating evidence.
The first one explains the shadow-AI numbers almost entirely. Reactance is the psychological response to feeling your autonomy is being restricted: when people are pushed, they push back, and warnings routinely function as recommendations. Berger’s remedy isn’t a better argument. It’s designing situations where people persuade themselves — offering a menu of options rather than a directive, asking rather than telling, and starting from an understanding of what the person is actually protecting. The fourth barrier, uncertainty, explains the rest of the gap: people stall when they can’t predict what a change will cost them, and the fix is making the new thing cheap and low-risk to try rather than proving it’s worth adopting wholesale.
Two ideas worth carrying into the boardroom: Diagnose the barrier before designing the intervention. Slow adoption looks the same from the top whether the cause is reactance, uncertainty, or an unusable tool — and the three require opposite responses. Pressure aimed at the wrong barrier doesn’t just fail; it manufactures the resistance it was meant to overcome. And lower the cost of trying instead of raising the cost of not trying. Mandates and compliance tracking raise the price of inaction, which produces documented adoption and very little real usage. Making something trivially easy to test, abandon, and return to produces the opposite.
Read The Catalyst →
Best for: Leaders who’ve already announced the change and can’t understand why it isn’t happening. Reading commitment: About five hours, brisk and example-driven.
Pendo is built to show exactly that: not whether people complied, but where they actually stopped. It’s a product and software experience platform built primarily for customer-facing products, but the same instrumentation works turned inward: point it at the internal tools employees are supposed to be adopting, and it records how they actually move through them — which features get opened, where a workflow gets abandoned, which steps quietly generate the most rework. On top of that picture it delivers in-app guidance at the moment someone hesitates, rather than in a training session three weeks earlier. It’s a repurposing, not the tool’s native use case — but it turns “our people aren’t adopting the tool” from an assertion into a specific location in a specific workflow.
It’s not a light purchase, and the trade-offs are well documented. Pricing is quote-based and enterprise-scaled, the learning curve is steep enough that reviewers regularly say the platform needs a dedicated owner, and its guidance layer attaches to interface elements in a way that requires maintenance whenever the underlying software changes. It also won’t tell you whether the tool deserved adoption in the first place. But for leaders who have been treating an adoption problem as a motivation problem, it usually reveals that the parking brake Berger describes is somewhere specific, boring, and fixable.
This issue also carries a placement opportunity: an exclusive feature in Entrepreneur, arranged through TEI’s media relationships. Entrepreneur reaches operators and founders who tend to judge a company by how it actually gets things adopted rather than by what it announced, which is exactly the distinction this issue is about. TEI’s team handles the interview, drafting, and placement end to end.
Curious? Reply here or book a call on our Calendly and our team will fill you in.
Take with you: If a third of your people would hide how they work from you, what is the mandate actually measuring?
The Executive Insight — 169 Madison Ave STE 11570, New York, NY, 10016
